用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt;运行前请先检查来源。
npx skills add https://github.com/diegosouzapw/awesome-omni-skill --skill ruler命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地?可下载 SkillsMP 当前能够提供的文件。
Token-efficient tracking for AI orchestration. CLI-first for status updates (~50 tokens), agent fallback for complex ops (~1KB). Use when: updating task status, querying blockers, creating progress files, validating phases.
AshAi extension guidelines for integrating AI capabilities with Ash Framework. Use when implementing vectorization/embeddings, exposing Ash actions as LLM tools, creating prompt-backed actions, or setting up MCP servers. Covers semantic search, LangChain integration, and structured outputs.
This skill should be used when solving hard questions, complex architectural problems, or debugging issues that benefit from GPT-5 Pro or GPT-5.1 thinking models with large file context. Use when standard Claude analysis needs deeper reasoning or extended context windows.
正在显示 SKILL.md
基于 SOC 职业分类
| name | ruler |
| description | Unified AI agent configuration propagation across 18+ coding assistants. |
| version | 1.0.0 |
Unified AI agent configuration propagation across 18+ coding assistants.
Repository: https://github.com/intellectronica/ruler Documentation: https://deepwiki.com/intellectronica/ruler
Ruler centralizes AI agent instructions in .ruler/ and distributes them to all configured agents via ruler apply. Supports Model Context Protocol (MCP) propagation with merge/overwrite strategies.
.ruler/
├── *.md # Rules (concatenated alphabetically)
├── ruler.toml # Agent config + MCP settings
└── mcp.json # Shared MCP servers
npm install -g ruler
# or
npx ruler init
Creates .ruler/ directory with default files:
ruler init # Local project
ruler init --global # ~/.config/ruler/
Creates:
instructions.md - Central AI instructionsruler.toml - Configuration filemcp.json - MCP server definitionsPropagates rules to all configured agents:
ruler apply # All default agents
ruler apply --agents claude,codex # Specific agents
ruler apply --no-mcp # Skip MCP propagation
ruler apply --mcp-overwrite # Replace native MCP configs
ruler apply --no-gitignore # Skip .gitignore updates
Restores files from backups:
ruler revert # Restore all, delete backups
ruler revert --keep-backups # Restore but keep .bak files
| Agent | Identifier | Instructions Output | MCP Config |
|---|---|---|---|
| GitHub Copilot | copilot | .github/copilot-instructions.md | .vscode/mcp.json |
| Claude Code | claude | CLAUDE.md | .mcp.json |
| OpenAI Codex CLI | codex | AGENTS.md | .codex/config.toml |
| Jules | jules | AGENTS.md | - |
| Cursor | cursor | .cursor/rules/ruler_cursor_instructions.mdc | .cursor/mcp.json |
| Windsurf | windsurf | .windsurf/rules/ruler_windsurf_instructions.md | ~/.codeium/windsurf/mcp_config.json |
| Cline | cline | .clinerules | - |
| Amp | amp | AGENT.md | - |
| Aider | aider | ruler_aider_instructions.md + .aider.conf.yml | .mcp.json |
| Firebase Studio | firebase | .idx/airules.md | - |
| Open Hands | openhands | .openhands/microagents/repo.md | .openhands/config.toml |
| Gemini CLI | gemini-cli | GEMINI.md |
Complete configuration reference:
# ═══════════════════════════════════════════════════════════════════
# DEFAULT AGENTS
# ═══════════════════════════════════════════════════════════════════
# Agents to run when --agents flag is not specified
# If omitted, all agents are active
default_agents = ["copilot", "claude", "codex", "cursor", "amp"]
# ═══════════════════════════════════════════════════════════════════
# GLOBAL MCP CONFIGURATION
# ═══════════════════════════════════════════════════════════════════
[mcp]
# Enable/disable MCP propagation globally (default: true)
enabled = true
# How to combine with native configs: "merge" or "overwrite"
# merge: Union of servers, incoming takes precedence on conflicts
# overwrite: Replace native config entirely
merge_strategy = "merge"
# ═══════════════════════════════════════════════════════════════════
# GITIGNORE CONFIGURATION
# ═══════════════════════════════════════════════════════════════════
[gitignore]
# Auto-add generated files to .gitignore (default: true)
enabled = true
# ═══════════════════════════════════════════════════════════════════
# AGENT-SPECIFIC CONFIGURATIONS
# ═══════════════════════════════════════════════════════════════════
[agents.copilot]
enabled = true
output_path = ".github/copilot-instructions.md"
[agents.claude]
enabled =
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
Define shared MCP servers:
{
"mcpServers": {
"gay": {
"type": "stdio",
"command": "julia",
"args": ["--project=@gay", "-e", "using Gay; Gay.serve_mcp()"],
"env": {
"GAY_SEED": "1069"
}
},
"firecrawl": {
"type": "stdio",
"command": "npx",
"args": ["-y", "firecrawl-mcp"],
"env": {
"FIRECRAWL_API_KEY":
All agents implement:
interface IAgent {
getIdentifier(): string; // e.g., "copilot", "claude"
getName(): string; // Human-readable name
getDefaultOutputPath(): string | Record<string, string>;
getMcpServerKey?(): string; // e.g., "servers" for Copilot
applyRulerConfig(
concatenatedRules: string,
projectRoot: string,
rulerMcpJson: object,
agentConfig?: AgentConfig
): Promise<void>;
}
| Agent | MCP Server Key |
|---|---|
| Copilot | servers |
| Claude | mcpServers |
| OpenCode | mcp |
| Others | mcpServers |
1. Load ruler.toml configuration
2. Find .ruler/ directory (local or global)
3. Concatenate all *.md files alphabetically
4. Load and validate mcp.json
5. For each selected agent:
a. Determine output paths (with overrides)
b. Backup existing files (.bak)
c. Call agent.applyRulerConfig()
d. If MCP enabled:
- Read native MCP config
- Merge with ruler mcp.json
- Write merged config
6. Update .gitignore with generated paths
Combines servers from both configs, incoming takes precedence:
// base (native)
{ "mcpServers": { "a": {...}, "b": {...} } }
// incoming (ruler)
{ "mcpServers": { "b": {...}, "c": {...} } }
// result
{ "mcpServers": { "a": {...}, "b": {...from ruler}, "c": {...} } }
Replaces native config entirely:
// result = incoming config only
{ "mcpServers": { "b": {...}, "c": {...} } }
file.bak.bak, then delete .bak--keep-backups: restore but keep .bak files.openhands/config.tomlstdio_servers key (TOML format).vscode/settings.jsonopencode.jsonmcp key instead of mcpServers.mdc extension for instructions.aider.conf.ymlCLI flags (highest)
↓
ruler.toml settings
↓
Built-in defaults (lowest)
Examples:
--no-gitignore overrides [gitignore] enabled = true--agents claude overrides default_agents[agents.X.mcp] overrides global [mcp]# 1. Initialize
ruler init
# 2. Add rules
cat > .ruler/01-style.md << 'EOF'
# Code Style
- Use TypeScript
- Prefer functional programming
- Maximum line length: 100
EOF
cat > .ruler/02-security.md << 'EOF'
# Security
- Never log secrets
- Use parameterized queries
- Validate all inputs
EOF
# 3. Configure MCP
cat > .ruler/mcp.json << 'EOF'
{
"mcpServers": {
"gay": {
"command": "julia",
"args": ["--project=@gay", "-e", "using Gay; Gay.serve_mcp()"]
}
}
}
EOF
# 4. Configure agents
cat > .ruler/ruler.toml << 'EOF'
default_agents = ["claude", "codex", "amp"]
[mcp]
enabled = true
merge_strategy = "merge"
[agents.claude]
enabled = true
[agents.codex]
enabled = true
[agents.amp]
enabled = true
EOF
# 5. Apply
ruler apply
# 6. Verify
cat CLAUDE.md
cat AGENTS.md
cat AGENT.md
# 7. Iterate and refine
vim .ruler/01-style.md
ruler apply
# 8. Revert if needed
ruler revert
The .ruler/ directory in music-topos extends the standard ruler with:
[bisimulation]
enabled = true
polarity_rotation = true
[bisimulation.agents]
claude = "PLUS" # +1
codex = "ERGODIC" # 0
cursor = "MINUS" # -1
# Sum = 0 mod 3 ✓
(def AGENTS
{:codex {:trit 0 :skills-path ".codex/skills"}
:claude {:trit -1 :skills-path ".claude/skills"}
:amp {:trit 0 :skills-path ".ruler/skills"}
:cursor {:trit -1 :skills-path ".cursor/skills"}
:copilot {:trit 1 :skills-path ".vscode/skills"}
:aider {:trit 1 :skills-path ".skillz"}})
;; Propagate with GF(3) conservation check
(propagate-all!)
This skill connects to the K-Dense-AI/claude-scientific-skills ecosystem:
general: 734 citations in bib.duckdbThis skill connects to Software Design for Flexibility (Hanson & Sussman, 2021):
Concepts: autonomous agent, game, synthesis
ruler (+) + SDF.Ch10 (+) + [balancer] (+) = 0
Skill Trit: 1 (PLUS - generation)
Adventure games synthesize techniques. This skill integrates multiple patterns.
This skill maps to Cat# = Comod(P) as a bicomodule in the equipment structure:
Trit: 0 (ERGODIC)
Home: Prof
Poly Op: ⊗
Kan Role: Adj
Color: #26D826
The skill participates in triads satisfying:
(-1) + (0) + (+1) ≡ 0 (mod 3)
This ensures compositional coherence in the Cat# equipment structure.
.gemini/settings.json |
| Junie | junie | .junie/guidelines.md | - |
| AugmentCode | augmentcode | .augment/rules/ruler_augment_instructions.md | .vscode/settings.json |
| Kilo Code | kilocode | .kilocode/rules/ruler_kilocode_instructions.md | .kilocode/mcp.json |
| OpenCode | opencode | AGENTS.md | opencode.json |
| Goose | goose | .goosehints | - |
| Crush | crush | CRUSH.md | .crush.json |